ORIGINAL ARTICLE
Social Distance, Framing, and Judgment: AConstrual Level Perspective
Xiaoli Nan
Department of Life Sciences Communication, University of Wisconsin–Madison, WI 53706
Drawing upon construal level theory, this research investigates the influence of social
distance on individuals’ responses to persuasive messages. Experiment 1 (N = 133)
demonstrates that the persuasive impact of a gain frame becomes stronger when people
make judgments for socially distant (e.g., others) versus proximal entities (e.g., selves).
On the other hand, the persuasive impact of a loss frame remains the same across dif-
ferent levels of social distance. Experiment 2 (N = 135) shows that the persuasiveness
of a societal frame becomes stronger when people make judgments for socially distant
versus proximal entities, whereas the persuasiveness of an individual frame is unaf-
fected by social distance. Experiment 3 (N = 80) provides evidence that mental salience
of positive and societal outcomes of an action increases as social distance increases,
whereas mental salience of negative and individual outcomes remains the same across
different levels of social distance.
doi:10.1111/j.1468-2958.2007.00309.x
A tree, when viewed from a far distance, may simply appear to be a pile of greenleaves. When observing closely, however, you may see more details: that it has
exceptionally large leaves, that it has thick branches, and that it hosts many birdnests. Just as physical distance may influence how we perceive the world, psycho-
logical distance, too, can have an impact on our perceptions. Construal level theory(CLT) (Trope & Liberman, 2003) postulates that people construct different mental
representations of the same information depending on whether the informationpertains to the near future (i.e., psychologically proximal) or the distant future(i.e., psychologically distant). For example, in one study, Liberman and Trope
(1998) asked participants to imagine themselves watching TV either tomorrow ornext year and describe this activity. Interestingly, those asked to imagine watching
TV tomorrow came up with concrete descriptions such as ‘‘sitting on the sofa,
Corresponding author: Xiaoli Nan; e-mail: [email protected] article was accepted under the editorship of Jim Dillard. An earlier version of this articlereceived Top Faculty Paper Award from the Communication Theory and MethodologyDivision of the 2006 AEJMC conference.
Human Communication Research ISSN 0360-3989
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 489
flipping channels,’’ whereas those asked to imagine watching TV next year thoughtof abstract descriptions such as ‘‘being entertained.’’
According to Trope and his colleagues, the principles of CLT apply to variousforms of psychological distance, including the so-called social distance (e.g., self vs.
other, in-group vs. out-group). People may construct different mental representa-tions of the same information depending on whether the information pertains toselves or others. In light of CLT, the current research seeks to understand the impact
of social distance on individuals’ responses to persuasive messages utilizing variousframes. Specifically, this research examines whether the persuasive impact of various
message frames (i.e., gain frames, loss frames, societal frames, individual frames)differs as a function of the reference group (e.g., selves vs. others) for whom judg-
ments are made after message exposure.
Theoretical background
Construal Level Theory
A basic tenet of CLT (Trope & Liberman, 2003) is that temporal distance influencesthe way people mentally represent the same information or event. The greater the
temporal distance, the more likely are events to be represented in terms of abstract,schematic, and decontextualized features, often referred to as high-level construals.
On the other hand, as temporal distance decreases, events are more likely to berepresented in terms of concrete, detailed, and contextualized features, often referred
to as low-level construals.In support of these propositions, Liberman and Trope (1998) found that stu-
dents, when asked to imagine themselves engaging in several activities (e.g., movinginto a new apartment) either tomorrow or next year and to describe these activities,generated more high-level construals (e.g., starting a new life) in the distant future
condition than in the near future condition. In contrast, more low-level construals(e.g., packing and carrying boxes) were generated in the near future condition
than in the distant future condition. In a series of studies, Forster, Friedman, andLiberman (2004) demonstrate that a distant future time perspective facilitates
abstract thinking, whereas a near future time perspective facilitates concrete think-ing. They found that participants who envisioned their lives and imagined them-
selves engaging in a task 1 year later as opposed to the next day subsequentlyperformed better on a series of insight tasks requiring abstract thinking. On theother hand, a near future time perspective improved analytical problem-solving skills
compared to a distant future time perspective.Temporal distance is one dimension of what Lewin (1951) called psychological
distance, which also includes the so-called social distance (e.g., self vs. other, in-groupvs. out-group). Although CLT has primarily dealt with temporal distance, CLT can
be considered as a unified theory of psychological distance (Trope & Liberman, 2003,p. 403). Thus, we should expect that social distance, like temporal distance, would
influence the way people mentally represent the same information or event.
Social Distance, Framing, and Judgment X. Nan
490 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
Social distance as a form of psychological distance
In light of CLT, as social distance increases, information will be represented in more
abstract, schematic, and decontextualized terms (i.e., high-level construals). On theother hand, as social distance decreases, information will be represented in more
concrete, detailed, and contextualized terms (i.e., low-level construals). Consistentwith these assumptions, previous research has shown that people tend to explainothers’ behaviors in dispositional (i.e., high-level, abstract) terms and their own
behavior in situational (i.e., low-level, concrete) terms (e.g., Fiedler, Semin,Finkenauer, & Berkel, 1995). Other research has demonstrated that people tend to
perceive in-groups and out-groups differently. Compared with in-groups, out-groups are perceived as more homogenous (e.g., Jones, Wood, & Quattrone,
1981), as having more predictable sets of properties (e.g., Linville, Fischer, & Yoon,1996), and are described in more abstract terms (e.g., Fiedler et al., 1995).
According to CLT, differences in mental representations of events associatedwith large versus small psychological distance can have significant evaluative con-
sequences. More specifically because high-level construals become more salient aspsychological distance increases, we should expect judgments about the event bemore determined by high-level construals when psychological distance is large versus
small. In contrast, low-level construals should be more influential when psycholog-ical distance is small versus large.
These propositions are supported by a number of studies (e.g., Ebert, 2005;Liberman & Trope, 1998; Pennington & Roese, 2003; Ziamou & Veryzer, 2005). For
example, Ziamou and Veryzer demonstrate that, in forming judgments about a newproduct, people put more weight on the product’s functionality (i.e., a high-level
construal) when the purchase is expected to occur in the distant versus near future.On the other hand, the product’s interface design (i.e., a low-level construal) is moreinfluential when the purchase is expected to occur in the near versus distant future.
More relevant to this research, Ebert (2005) investigates the influence of socialdistance on the way people make judgments. In one of her studies, participants were
asked to list the long-term benefits (i.e., high-level construals) and short-term costs(i.e., low-level construals) of a particular course of action. Then half of the partici-
pants rated the importance of the long-term benefits and short-term costs for them-selves, whereas the other half did the rating for their friends. Interestingly, Ebert
found that short-term costs were rated as being much more important when thejudgments were made for selves versus friends. Long-term benefits, on the other
hand, were perceived as equally important under the two conditions. Consistent withCLT, Ebert’s research suggests that low-level construals such as short-term costs ofa course of action tend to be more influential in forming judgments for a socially
proximal entity (e.g., self) than for a socially distant entity (e.g., a friend).
Gain–loss framing and construal level
Gain–loss framing, sometimes referred to as message framing or goal framing (Levin,
Schneider, & Gaeth, 1998), has recently received considerable attention in the area of
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 491
persuasive communication (e.g., Block & Keller, 1995; Davis, 1995; Maheswaran &Meyers-Levy, 1990; Meyerowitz & Chaiken, 1987; Mitchell, 2001; Rothman, Bartels,
Wlaschin, & Salovey, 2006; Umphrey, 2003; for reviews, see O’Keefe & Jensen, 2006,in press-a, in press-b). Rooted in prospect theory (Kahneman & Tversky, 1979),
research on gain–loss framing has mostly been concerned about the relative persua-siveness of a gain versus loss frame. In a persuasive message, a gain frame emphasizesthe positive outcomes of compliance (e.g., ‘‘If you apply sunscreen, your skin will be
protected from the sun’s relentless ultraviolet radiation.’’), whereas a loss framehighlights the negative outcomes of noncompliance (e.g., ‘‘If you don’t apply sun-
screen, your skin will be harmed by the sun’s relentless ultraviolet radiation.’’).A primary finding from previous research is that a loss frame tends to be more
persuasive than a gain frame (e.g., Meyerowitz & Chaiken, 1987; Rothman, Salovey,Antone, Keough, & Martin, 1993), although reversed as well as null effects have also
been reported (e.g., Block & Keller, 1995; Maheswaran & Meyers-Levy, 1990).Rather than focusing on the relative persuasiveness of a gain versus loss frame, this
research approaches gain–loss framing from the perspective of CLT, examining how
the effectiveness of each separate frame differs as a function of social distance. Recentstudies suggest that a gain frame may differ from a loss frame in terms of construal
level (e.g., Eyal, Liberman, & Trope, 2004; Pennington & Roese, 2003). For example,Eyal et al. (2004) speculate that positive attributes associated with a course of action
may constitute a higher construal level than negative attributes. The researchers pro-vided support for this posit by demonstrating the subordination of negative attributes
(or cons) to positive attributes (or pros), which means that the importance of negativeattributes is dependent on the value of positive attributes more than the importance of
the positive attributes is dependent on the value of the negative attributes (the notionof asymmetric conditional importance). In an experiment utilizing a 2 3 2 factorialdesign, Eyal et al. had college students evaluate the attractiveness of a student loan
offer. Half of the participants were told by a specialist that the student loan had eithersome pros or no pros. They were further asked to indicate their interest in knowing if
the loan had any cons. The other half was informed that the loan had either some consor no cons. They were then asked to indicate their interest in knowing if the loan had
any pros. It turned out that participants’ interest in knowing more about the cons ofthe student loan was affected by the initial information about the pros: Interest was
considerably lower when there were no pros than when there were some pros. On theother hand, participants’ interest in knowing more about the pros of the student loanwas not affected by the initial information about the cons: Interest remained high
regardless of whether there were no cons or some cons. These findings demonstratedthe subordination of cons to pros, that is, the importance of cons depends on whether
there are pros more than the importance of pros depends on whether there are cons.Eyal et al. concluded that positive attributes associated with a course of action con-
stitute a higher construal level than negative attributes.Further supporting the assertion that positive attributes constitute a higher con-
strual level than negative attributes, Eyal et al. found that people generated more
Social Distance, Framing, and Judgment X. Nan
492 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
arguments in favor of a social plan when it was expected to be launched in the distantversus near future. In comparison, more arguments against the social plan were
generated when the plan was expected to be launched in the near versus distantfuture. In a similar vein, Pennington and Roese (2003) suggest that positive or
promotional outcomes of an event constitute high-level construals, whereas negativeor preventional outcomes represent low-level construals. In one of their studies, theresearchers asked students to indicate their concerns about an upcoming exam at two
time points: 2 weeks before and on the day of the exam. The researchers found thatstudents were more concerned about positive or promotional outcomes (e.g., getting
a high score) when giving the ratings 2 weeks before versus on the day of the exam.In comparison, concerns for negative or preventional outcomes (e.g., disappointing
oneself) remained the same across the two time points.It is important to note that both the literature and the current research have
adopted a broad treatment of construals. By definition, construals should refer rep-resentations of information in people’s minds. Researchers have implicitly extendedthe range of construals to encompassing information that exists outside of people’s
minds when claiming that pros (or promotional outcomes) are high-level construalsand cons (or preventional outcomes) are low-level construals. Thus, construals are
used here to refer not only to representations of information in people’s minds butalso to information external to people’s minds (e.g., the content of a persuasive
message).In sum, the literature suggests that a gain frame, which focuses on the positive
outcomes of compliance with a persuasive message, is associated with a higher con-strual level than a loss frame, which emphasizes the negative outcomes of noncom-
pliance. Based on CLT, I expected that a gain frame would be more persuasive whenjudgments are made for socially distant entities (e.g., others) than for socially prox-imal entities (e.g., selves). I also expected that a loss frame would be more persuasive
when judgments are made for socially proximal entities than for socially distantentities. These propositions are summarized herein.
H1: There will be an interaction between social distance and gain–loss framing such that
(a) the persuasive impact of a gain frame will increase when people make judgments
for a socially distant entity versus a socially proximal entity and (b) the persuasive
impact of a loss frame will decrease when people make judgments for a socially distant
entity versus a socially proximal entity.
Societal–individual framing and construal level
Societal–individual framing has received less research attention compared to gain–
loss framing. A societal frame is defined here as a message strategy that focuses on theimplications of complying or not complying with the advocacy for the society in
general (e.g., Taking public transit instead of driving a car for daily commute pro-vides cleaner air for people in your community.). An individual frame, on the other
hand, emphasizes the consequences of compliance or noncompliance for the
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 493
individual (e.g., Taking public transit instead of driving a car for daily commuteprovides cleaner air for you.).
In general, societal implications seem relatively abstract and decontextualized,whereas personal consequences appear fairly concrete and contextualized. Thus, we
may expect that a societal frame be generally associated with a higher construal levelthan an individual frame. Although no direct evidence is available to support thisassertion, recent work by Shah, Kwak, Schmierbach, and Zubric (2004) provides
indirect support for the validity of this proposition. Part of Shah et al.’s research wasto investigate the relative effect of a societal versus individual frame in a news story
on individuals’ cognitive complexity. The researchers found that participants gen-erated a more detailed description of the attributes of the central issue after being
exposed to a message framed at the individual level than they did after being exposedto a message framed at the societal level. Assuming what the participants thought
about was a reflection of the message content, it could be argued that an individualframe was more concrete and detailed than a societal frame and thus was associatedwith a lower construal level.
Based on the above arguments, I expected that a social frame would be morepersuasive when people make judgments for socially distant entities compared to
when judgments are made for socially proximal entities. In addition, I expected thatan individual frame would be more persuasive when people make judgments
for socially proximal versus distant entities. Thus, the following hypotheses areproposed.
H2: There will be an interaction between social distance and societal–individual framing
such that (a) the persuasive impact of a societal frame will increase when people make
judgments for a socially distant entity versus a socially proximal entity and (b) the
persuasive impact of an individual frame will decrease when people make judgments
for a socially distant entity versus a socially proximal entity.
The hypotheses were tested in three experiments. The purpose of Experiment 1 was
to test H1 that predicted the differential persuasive effect of a gain frame (H1a) andthat of a loss frame at different levels of social distance (H1b). The second experi-
ment was conducted to test H2 that posited the differential persuasive effect ofa societal frame (H2a) and that of an individual frame at different levels of social
distance (H2b). Whereas Experiments 1 and 2 tested the hypotheses using persuasivemessages as stimuli, Experiment 3 utilized a thought-listing technique to test the keyassumptions underlying the hypotheses.
Experiment 1
Participants and procedure
One hundred and thirty three undergraduate students recruited from introductorycommunication classes at the University of Wisconsin—Madison participated in the
experiment in exchange for extra credits. The experiment was conducted in a research
Social Distance, Framing, and Judgment X. Nan
494 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
lab, where students participated in the study in groups. The experiment utilized a 2(framing: gain vs. loss frame) 3 2 (social distance: proximal vs. distant) mixed
design, where framing was a between-subjects factor and social distance wasa within-subjects factor. Participants were randomly assigned to one of the two
framing conditions.Upon arrival at the research lab, participants were given a booklet, which con-
tained a cover story, a persuasive message in the form of a public service advertise-
ment (PSA), and a battery of questions probing message effects, perceived healthrisk, and brief demographics. The cover story on the first page indicated that the
purpose of the study was to understand college students’ responses to a health-related PSA. Participants were instructed to view the PSA on the next page at their
own pace. After viewing the ad, participants responded to the questionnaire attachedat the end of the booklet. Participants were debriefed and thanked after they returned
the booklet.
Message topic
The PSA encouraged people to take a hepatitis C test. Hepatitis C is a liver diseasecaused by the hepatitis C virus (HCV). The infection is typically spread by contact
with the blood of an infected person. Although Hepatitis C is not an uncommondisease—almost 4 million Americans are inflicted with it—there have been relatively
few health campaigns about preventing or detecting the disease. The PSA featureda headline and then described the potential risk of contracting HCV. Following this
information, an argument was made about the consequence of either taking or nottaking a hepatitis C test (see Table 1 for a description of the PSA).
Manipulations
Gain–loss framing was manipulated through different wording of the argument in
the PSA. In the gain frame, the argument emphasized the positive outcome of takinga hepatitis C test (i.e., ‘‘Remember, if you take a hepatitis C test in time, you may
enjoy the benefits of affordable treatment options for early-detected virus.’’). In theloss frame, the argument focused on the negative outcome of not taking a hepatitis C
test (i.e., ‘‘Remember, if you do not take a hepatitis C test in time, you may suffer thecosts of expensive treatment options for late-detected virus.’’). Similar manipula-
tions of gain–loss framing have been used in previous studies (e.g., Block & Keller,1995; Rothman et al., 1993).
Social distance was manipulated in the questionnaire. In the proximal condition,
participants were asked to make judgments about hepatitis C tests from the perspec-tive of their best friend. In the distant condition, participants were asked to make
judgments from the perspective of an average undergraduate student.
Measures
The key dependent measure was issue judgment regarding taking a hepatitis C test.
Following the prompt in the proximal social distance condition (i.e., ‘‘What would
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 495
Table
1D
escr
ipti
on
so
fth
eP
SAs
PSA
Des
crip
tio
n
Hep
atit
isC
Tes
tP
rom
oti
on
(Exp
erim
ent
1)
Th
isp
rin
tP
SAw
ascr
eate
db
ased
on
ah
epat
itis
Cin
form
atio
nal
bro
chu
red
istr
ibu
ted
on
lin
eb
yth
eC
DC
.
Th
eP
SAu
sed
the
sam
ep
ictu
re(t
he
imag
eo
fth
ree
ind
ivid
ual
s:a
you
ng
mal
e,a
mid
dle
aged
mal
e,an
da
mid
dle
aged
fem
ale)
that
was
feat
ure
din
the
CD
Cb
roch
ure
.T
he
hea
dli
ne
of
the
PSA
was
‘‘Tak
e
AH
epat
itis
CT
est.
’’F
oll
ow
ing
this
hea
dli
ne
was
ap
arag
rap
hd
escr
ibin
gth
eri
sko
fh
epat
itis
C:
‘‘Hep
atit
isC
isa
live
rd
isea
seca
use
db
yth
eh
epat
itis
Cvi
rus
(HC
V),
wh
ich
isfo
un
din
the
blo
od
of
per
son
sw
ho
hav
eth
isd
isea
se.
Un
kno
wn
tom
any
peo
ple
,H
epat
itis
Cis
hig
hly
con
tagi
ou
s.’’
Th
en,
an
argu
men
tw
asm
ade
abo
ut
the
con
seq
uen
ceo
fei
ther
taki
ng
(‘‘R
emem
ber
,if
you
take
ah
epat
itis
Cte
st
inti
me,
you
may
enjo
yth
eb
enef
its
of
affo
rdab
letr
eatm
ent
op
tio
ns
for
earl
y-d
etec
ted
viru
s.’’)
or
no
t
taki
ng
ah
epat
itis
Cte
st(‘
‘Rem
emb
er,
ifyo
ud
on
ot
take
ah
epat
itis
Cte
stin
tim
e,yo
um
aysu
ffer
the
cost
so
fex
pen
sive
trea
tmen
to
pti
on
sfo
rla
te-d
etec
ted
viru
s.’’)
.In
the
end
,th
eP
SAin
vite
dvi
ewer
sto
visi
tth
eC
DC
Web
site
for
mo
rein
form
atio
nab
ou
th
epat
itis
C.
Pu
bli
cT
ran
sit
Pro
mo
tio
n
(Exp
erim
ent
2)
Th
isp
rin
tP
SAw
ascr
eate
db
ased
on
air
po
llu
tio
nin
form
atio
no
nth
eE
nvi
ron
men
tal
Pro
tect
ion
Age
ncy
Web
site
,ac
cord
ing
tow
hic
hca
rsar
eo
ne
of
the
top
con
trib
uto
rsto
air
po
llu
tio
n.
Th
eP
SAfe
atu
red
a
pic
ture
sho
win
ga
wid
eci
tyst
reet
pac
ked
wit
hca
rsan
do
ther
mo
tor
veh
icle
s.A
irco
nta
min
atio
nca
nb
e
easi
lyse
enfr
om
the
clo
ud
ines
so
fth
eai
rab
ove
the
veh
icle
s.R
igh
tb
elo
wth
ep
ictu
rew
asan
argu
men
t
rega
rdin
gth
eb
enef
its
of
taki
ng
pu
bli
ctr
ansi
tth
atem
ph
asiz
edth
eh
ealt
hb
enef
its
of
taki
ng
pu
bli
ctr
ansi
t
eith
erfo
rth
eso
ciet
yin
gen
eral
(‘‘C
han
ces
are
peo
ple
inyo
ur
com
mu
nit
yw
ill
hav
eh
ealt
hie
rlu
ngs
ifyo
u
take
ap
ub
lic
tran
spo
rtat
ion
inst
ead
of
keep
ing
dri
vin
gyo
ur
ow
nca
r!’’)
or
for
the
ind
ivid
ual
(‘‘C
han
ces
are
you
wil
lh
ave
ah
ealt
hie
rlu
ng
ifyo
uta
kea
pu
bli
ctr
ansp
ort
atio
nin
stea
do
fke
epin
gd
rivi
ng
you
ro
wn
car!
’’).
Th
isp
arag
rap
hth
enap
pea
rsri
ght
bel
ow
the
argu
men
t:‘‘A
cco
rdin
gto
the
U.S
.E
nvi
ron
men
tal
Pro
tect
ion
Age
ncy
(EP
A),
dri
vin
ga
car
isth
esi
ngl
em
ost
po
llu
tin
gth
ing
that
mo
sto
fu
sd
o.
Cle
anai
ris
crit
ical
for
hea
lth
ylu
ngs
.’’
Note:
CD
C=
Cen
ters
for
Dis
ease
Co
ntr
ol
and
Pre
ven
tio
n;
PSA
=p
ub
lic
serv
ice
adve
rtis
emen
t.
Social Distance, Framing, and Judgment X. Nan
496 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
you tell your best friend about hepatitis C tests?’’) were two items tapping into issuejudgment. The first was ‘‘that he/she should take a hepatitis C test soon.’’ The second
was ‘‘that he/she should take a hepatitis C test in the near future.’’ Participants wereasked to indicate their agreement with each statement on a 1 (strongly disagree) to
7 (strongly agree) scale. The same two statements followed the prompt in the distantcondition (i.e., ‘‘What would you tell an average undergraduate student about hep-atitis C tests?’’). The two items were averaged to form an index for issue judgment
(a = .97 across social distance conditions).Participants’ perceived risk of contracting HCV was also measured. Similar to
issue judgment, perceived risk was measured at the two levels of social distance. First,participants were asked to indicate the likelihood that their best friend would con-
tract HCV on a 1 (strongly disagree) to 7 (strongly agree) scale. Then, the samequestion was asked with regard to an average undergraduate student.
Results
Hypothesis testing
H1 predicted an interaction between social distance and gain–loss framing. Specifi-cally, H1a proposed that a gain frame would be more persuasive when judgmentswere made for a socially distant versus proximal entity, whereas H1b posited that the
reverse would be true for a loss frame. To test H1a and H1b, a mixed analysis ofvariance (ANOVA) was performed, where gain–loss framing served as the between-
subjects factor and social distance was the within-subjects factor. The dependentvariable was issue judgment.
The analysis revealed a number of significant effects. First, there was a significantmain effect of social distance, F(1, 131) = 8.48, p, .01, h2 = .05. Issue judgment was
more positive in the distant condition (M = 3.58, SD = 1.60) than in the proximalcondition (M = 3.34, SD = 1.50), indicating that participants in general thought an
average undergraduate student should take a hepatitis C test, more so than their bestfriend. More relevant to the hypotheses, a significant interaction between socialdistance and gain–loss framing was detected, F(1, 131) = 7.51, p , .01, h2 = .05
(see Figure 1). Additional analyses revealed that the gain frame led to a significantlymore favorable issue judgment in the distant condition (M = 3.82, SD = 1.65) than in
the proximal condition, M = 3.34, SD = 1.58, F(1, 64) = 14.92, p , .001, h2 = .19.Thus, H1a was supported. On the other hand, no difference in issue judgment was
found for the loss frame across the two social distance conditions: distant: M = 3.35,SD = 1.52; proximal: M = 3.34, SD = 1.41, F(1, 67) , 1. H1b was not supported.
Power analysis indicated, however, that the design might fail to detect a small effect(power = .21 assuming f 2 = .02 or small effect size, power = .88 assuming f 2 = .15 ormedium effect size).
On a minor note, the main effect of gain–loss framing was not significant (F, 1),which seemed to contradict to prospect theory’s posit that losses loom larger than
gains (Kahneman & Tversky, 1979). However, the finding is not entirely surprising
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 497
given that the differential persuasive effect of gain- versus loss-framed messages isrelatively unstable in health contexts. Some studies did not observe such an effect(Rothman et al., 2006), perhaps due to the complexity of factors determining
people’s intention to enact health behaviors.
Exploring an alternative explanation
Previous research has shown that people tend to believe they are less susceptible tohealth risks than others (i.e., the self-positivity bias; Weinstein, 1987). This perceptual
bias could also emerge in situations involving two parties that hold different socialdistance to the self. Specifically, one may perceive less vulnerability to diseases for
a socially proximal entity (e.g., one’s best friend) than for a socially distant entity(e.g., a stranger). In this experiment, participants might hold a similar self-positivity
bias. In other words, participants might perceive less risk of contracting HCV fortheir best friend than for an average undergraduate student. If so, the differentialperceived risk could account for the interaction between social distance and gain–
loss framing and the findings supporting H1a.To examine whether self-positivity bias indeed occurred in the experiment,
a mixed ANOVA was performed. Gain–loss framing served as the between-subjectsfactor and social distance was the within-subjects factor. The dependent variable was
perceived risk. The results revealed only one strong main effect of social distance,F(1, 131) = 99.89, p, .001, h2 = .43. Perceived risk was much greater for the socially
distant entity (i.e., an average undergraduate student; M = 3.44, SD = 1.19) than forthe socially proximal entity (i.e., one’s best friend; M = 2.57, SD = 1.34).
To explore whether self-positivity bias accounted for the observed interaction
between social distance and gain–loss framing, and the findings supporting H1a,
3
3.2
3.4
3.6
3.8
4
Proximal DistantSocial Distance
Gain FrameLoss Frame
Figure 1 Persuasion as a function of social distance and gain-loss framing.
Social Distance, Framing, and Judgment X. Nan
498 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
a mixed analysis of covariance was conducted. Gain–loss framing served as thebetween-subjects factor, social distance was the within-subjects factor, and perceived
risk for one’s best friend and for an average undergraduate student served as twocovariates. The dependent variable was issue judgment. When the two risk percep-
tions were controlled for, the previous main effect of social distance on issue judg-ment disappeared (F, 1), indicating that perceived risk mediated the main effect ofsocial distance on issue judgment. However, the interaction between social distance
and gain–loss framing remained significant, F(1, 129) = 5.90, p , .05, h2 = .04.Additional analyses (with the covariates controlled for) revealed that the gain frame
led to a significantly more positive issue judgment in the distant condition (M = 3.82,SE = .19) than in the proximal condition, M = 3.34, SE = .18, F(1, 62) = 14.72,
p , .001, h2 = .19. No difference in issue judgment was found for the loss frameacross the two social distance conditions. Overall, the results suggest that the differ-
ential perceived risk associated with different social distance, or self-positivity bias,does not account for the observed interactive effect of social distance and gain–lossframing on issue judgment. These findings lend further support to the validity of the
construal level explanation.
Discussion
This experiment found the expected interactive effect of social distance and gain–loss
framing on persuasion. The interaction appeared to be primarily driven by a signifi-cant difference in judgments for the gain frame across the two social distance con-
ditions. Specifically, under the gain frame, judgments made for a socially distantentity were more positive, compared to the same judgments made for a socially
proximal entity. This finding is consistent with CLT’s proposition that judgmentsare more susceptible to the influence of high-level construals as psychological dis-tance increases. Further attesting to the viability of the construal level explanation,
additional evidence indicated that the main findings could not be attributed to analternative explanation involving self-positivity bias. Although perceived risk of
contracting the disease was found to be higher for the socially distant than for thesocially proximal, this effect did not account for the interaction between social
distance and gain–loss framing. It is worthy noting that perceived risk is just onedimension involved in self-positivity bias, which also includes less perceived severity
of negative events when they occur to selves versus others (Shah, Faber, & Youn,1999). Whether or not perceived severity could account for the observed interactionneeds to be addressed in future studies.
This experiment did not find a differential effect of the loss frame on judgmentsassociated with different levels of social distance. Power analysis showed that, given
the current design, the experiment might miss small effects. Thus, methodologicalimprovements such as increasing sample size may prove helpful in detecting the
effects. On the other hand, some recent studies applying CLT indicate that low-levelconstruals may be equally salient at different levels of psychological distance. For
instance, Pennington and Roese (2003) demonstrated that people’s promotional
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 499
concerns (i.e., concerns about the positive outcomes of an event), which representhigh-level construals, increase as temporal distance increases. In contrast, people’s
preventional concerns (i.e., concerns about the negative outcomes of an event),which constitute low-level construals, do not vary with temporal distance. Therefore,
it could be that the salience of low-level construals is not as affected by psychologicaldistance, compared to the salience of high-level construals.
Regardless, a second experiment was conducted to provide a further test for the
overarching proposition that the weights assigned to low-level construals and high-level construals in judgments are both influenced by psychological distance. The
second experiment focused on societal–individual framing as a different operation-alization of construal level. As argued previously, a societal frame constitutes a high-
level construal about the advocated issue, whereas an individual frame representsa low-level construal. It was expected that a societal frame would be more persuasive
when judgments were made for a socially distant entity than for a socially proximalentity. The reverse effect was hypothesized for an individual frame.
Experiment 2
Participants and procedure
One hundred and thirty-five undergraduate students recruited from introductory
communication classes in the same university participated in this experiment inexchange for extra credits. The procedure was similar to that used in Experiment 1.
The experiment was conducted in a research lab, where students participated in thestudy in groups. It utilized a 2 (framing: societal vs. individual frame) 3 2 (social
distance: proximal vs. distant) mixed design, where framing was a between-subjectsfactor and social distance was a within-subjects factor. Participants were randomlyassigned to one of the two framing conditions.
Upon arrival at the research lab, participants were given a booklet, which con-tained a cover story, a persuasive message in the form of a PSA, and a battery of
questions probing message effects and brief demographics. The cover story on thefirst page indicated that the purpose of the study was to understand college students’
responses to a public transit PSA. Participants were instructed to view the PSA on thenext page at their own pace. After viewing the ad, participants responded to the
questionnaire attached at the end of the booklet. Participants were debriefed andthanked after they returned the booklet.
The PSA encouraged people to take public transit instead of driving their own
car. The ad presented an argument regarding the health benefits of taking publictransit. It was concluded with an Environmental Protection Agency message on car
driving and pollution (see Table 1 for a description of the PSA).
Manipulations
Societal–individual framing was manipulated through different wording of the argu-
ment in the PSA. In the societal frame, the argument was constructed such that it
Social Distance, Framing, and Judgment X. Nan
500 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
emphasized the health benefits of taking public transit for the society in general (i.e.,‘‘Chances are people in your community will have healthier lungs if you take a public
transportation instead of keeping driving your own car!’’). In the individual frame,the consequence of taking public transit was framed at an individual level (i.e.,
‘‘Chances are you will have a healthier lung if you take a public transportationinstead of keeping driving your own car!’’).
The manipulation of social distance was similar to that used in Experiment 1.
The only difference was that, in the proximal condition, participants were asked tomake judgments about taking public transit from their own perspective. In the
distant condition, they were asked to make judgments from the perspective of anaverage undergraduate student.
Dependent measure
Similar to Experiment 1, the key dependent measure was issue judgment regardingtaking public transit. Following the prompt in the distant condition (i.e., ‘‘Whatwould you tell an average undergraduate student about taking a public transporta-
tion?’’) were two items tapping into issue judgment. The first was ‘‘that he/she shouldtake a public transportation for his/her daily commute.’’ The second was ‘‘that taking
a public transportation for his/her daily commute is a good idea.’’ Participants wereasked to indicate their agreement with each statement on a 1 (strongly disagree) to
7 (strongly agree) scale. On two similar items, participants also made judgmentsabout taking public transit from their own perspective. The two items were averaged
to form an index for issue judgment (a = .87 across social distance conditions).
Results
H2 predicted an interaction between social distance and societal–individual framing.Specifically, H2a proposed that a societal frame would be more persuasive when
judgments were made for a socially distant versus proximal entity, whereas H2bposited that the reverse would be true for an individual frame. To test H2a and
H2b, a mixed ANOVA was performed, where societal–individual framing servedas the between-subjects factor and social distance was the within-subjects factor.
The dependent variable was issue judgment.The analysis revealed a number of significant effects. First, there was a significant
main effect of social distance, F(1, 133) = 5.41, p, .05, h2 = .04. Issue judgment wassignificantly more positive in the distant condition (M = 4.63, SD = 1.54) than in theproximal condition (M = 4.15, SD = 1.96). This finding indicates that participants in
general thought an average undergraduate student should take public transit, moreso than themselves. More relevant to the hypotheses, a significant interaction
between social distance and societal–individual framing was detected, F(1, 133) =4.36, p , .05, h2 = .03; see Figure 2. Additional analyses revealed that the societal
frame led to a significantly more favorable issue judgment in the distant condition(M = 4.84, SD = 1.47) than in the proximal condition, M = 3.97, SD = 1.95, F(1, 70)
= 9.48, p, .01, h2 = .11. Thus, H2a was supported. On the other hand, no difference
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 501
in issue judgment was found for the individual frame across the two social distance
conditions (distant: M = 4.39, SD = 1.60; proximal: M = 4.34, SD = 1.97, F, 1). H2bwas not supported. Power analysis indicated, however, that the design might fail to
detect a small effect (power = .20 assuming f 2 = .02 or small effect size, power = .86assuming f 2 = .15 or medium effect size).
Discussion
Overall, Experiment 2 produced findings that were consistent with those fromExperiment 1. An interactive effect of social distance and societal–individual framing
on persuasion was found. The societal frame was more persuasive when judgmentswere made for a socially distant versus proximal entity. These findings are congruentwith CLT, which predicts that the impact of high-level construals on judgments will
increase as psychological distance increases.This experiment did not find evidence for differential effect of the individual
frame across the two social distance conditions. The results of a power analysissuggested that the null effect might be due to insufficient statistical power to detect
small effects. Assuming the null effect was a reflection of a true null effect, the findingdoes not support CLT’s prediction regarding the differential judgmental effect of
low-level construals at different levels of psychological distance. However, it does notcontradict CLT. It is possible that whether or not psychological distance exerts aninfluence on the salience of low-level construals depends on factors such as
the operationalization of psychological distance (e.g., social distance or temporaldistance) and the operationalization of construal level (e.g., message framing or
3.8
4
4.2
4.4
4.6
4.8
5
Proximal DistantSocial Distance
Societal FrameIndividual Frame
Figure 2 Persuasion as a function of social distance and societal-individual framing.
Social Distance, Framing, and Judgment X. Nan
502 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
feasibility vs. desirability considerations). The same thing may be said for high-levelconstruals.
Both experiments provided compelling evidence that a gain frame and a societalframe used in persuasive messages tend to be more persuasive when judgments are
made for socially distant versus proximal entities. In contrast, the persuasiveness ofa loss frame and an individual frame is less dependent upon the level of socialdistance. A key assumption underlying the major hypotheses proposed here was
derived from CLT. That is, changes in social distance induce changes in people’smental representations of the same event. High-level construals become more salient
as social distance increases, whereas low-level construals become less salient as socialdistance increases. Based on findings of previous studies (e.g., Eyal et al., 2004;
Pennington & Roese, 2003; Shah et al., 2004), this research indicates that positiveand societal outcomes of an action constitute a higher construal level than negative
and individual outcomes of an action. Thus, the persuasiveness of a gain frame anda societal frame will increase as social distance increases, whereas the persuasivenessof a loss frame and an individual frame is expected to decrease as social distance
increases. However, an issue unaddressed in both experiments is the provisionof direct empirical evidence that social distance indeed influences the mental salience
of positive and societal outcomes of an action as well as negative and individualoutcomes of an action in a way that is consistent with the predictions of CLT.
A straightforward solution to providing such evidence is an experiment utilizinga thought-listing technique (cf. Eyal et al., 2004; Study 4). Hence, a third experiment
was conducted, the goals of which were to find out (a) whether mental salience ofpositive outcomes of an action increases as social distance increases, (b) whether
mental salience of negative outcomes of an action decreases as social distanceincreases, (c) whether mental salience of societal outcomes of an action increasesas social distance increases, and (d) whether mental salience of individual outcomes
of an action decreases as social distance increases.
Experiment 3
Participants and procedure
Eighty undergraduate students recruited from introductory communication classes in
the same university participated in the experiment in exchange for extra credits. Thefirst task in the experimental design was to select an action or behavior that is familiarto prospective research participants (i.e., college students). People will not have too
many thoughts about the outcomes of an unfamiliar action/behavior. The action/behavior should also be able to be linked with all four types of outcomes: positive,
negative, societal, and individual. Given these considerations, smoking was chosen asthe focal behavior. Potentially, there are two types of positive outcomes (i.e., the
positive outcomes of smoking and the positive outcomes of refraining from smoking)and two types of negative outcomes (i.e., the negative outcomes of smoking and the
negative outcomes of refraining from smoking) that can be explored. However, the
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 503
positive outcomes of refraining from smoking correspond to a gain frame in a per-suasive message and the negative outcomes of smoking correspond to a loss frame in
a persuasive message, and therefore, they are most relevant to the current research andto persuasive communication. (It would be odd to see a PSA emphasizing the positive
outcomes of smoking or the negative outcomes of refraining from smoking.)The experiment was conducted online. Participants were asked to list the positive
outcomes of refraining from smoking and the negative consequences of smoking.
They were instructed to put one outcome in one text box. Nine text boxes wereprovided, but the participants were told that they could list as many outcomes as
they wanted. The order in which participants listed positive and negative outcomeswas counterbalanced. That is, half of them listed positive outcomes first, whereas the
other half listed negative outcomes first. In addition, half of the participants gener-ated the outcomes from their own perspective, whereas the other half generated the
outcomes from the perspective of an average undergraduate student. Thus, theexperiment had a 2 (social distance: proximal vs. distant) 3 2 (order: positive firstvs. negative first) 3 2 (outcome valence: positive vs. negative) mixed design. Social
distance and order were between-subjects factors, whereas outcome valence wasa within-subjects factor.
Dependent variables
The first two dependent variables were the number of positive outcomes and thenumber of negative outcomes. Because the participants were asked to put one out-
come in one text box, the coding was straightforward. In a few instances thereappeared to be more than one outcome in one box (e.g., mouth disease/cancer,
increased risk of heart/lung problems). Two independent judges coded the numberof outcomes in each category and interjudge agreement was 95%.
The other dependent variables included the number of societal outcomes and the
number of individual outcomes. The participants were not explicitly asked to listsocietal and individual outcomes of smoking and refraining from smoking. However,
the expectation was that, given the nature of smoking, both societal and individualoutcomes would be generated and they might be under the category of either positive
or negative outcomes. This expectation was confirmed. Societal outcomes includedstatements such as, ‘‘(refraining from smoking) won’t expose friends/family to second
hand smoke,’’ ‘‘(smoking) puts others’ health in danger,’’ and ‘‘(smoking causes)environmental pollution.’’ Individual outcomes were thoughts such as, ‘‘(refrainingfrom smoking) helps keep your body in shape,’’ ‘‘(smoking leads to) yellow teeth,’’ and
‘‘(smoking results in) lung cancer.’’ Two independent judges coded the number ofsocietal and individual outcomes. Interjudge agreement was 86%.
Results
Two key assumptions underlying H1 was that mental salience of positive outcomes
of an action increases as social distance increases, whereas mental salience of negative
Social Distance, Framing, and Judgment X. Nan
504 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
outcomes decreases as social distance increases. To test these assumptions, a SocialDistance (proximal or distant) 3 Order (positive first or negative first) 3 Outcome
Valence (positive or negative) mixed ANOVA was performed on the number oflisted outcomes, with social distance and order as between-subjects factors and out-
come valence as a within-subjects factor.The results revealed a number of significant effects. First, the main effect of
outcome valence was significant, F(1, 76) = 43.82, p , .001, h2 = .26, indicating
that considerably more negative outcomes of smoking (M = 3.97, SD = 1.81) weregenerated than positive outcomes of refraining from smoking (M = 3.18, SD = 1.54).
Second, the interaction between outcome valence and order was significant, F(1, 76)= 23.42, p , .001, h2 = .14. The nature of the interaction was such that, when
positive outcomes were the first to be generated, negative outcomes (M = 3.86,SD = 1.79) moderately outnumbered positive outcomes, M = 3.60, SD = 1.72,
F(1, 42) = 3.92, p = .05, h2 = .09. On the other hand, when negative outcomes werethe first to be generated, the number of negative outcomes (M = 4.11, SD = 1.84)considerably exceeded the number of positive outcomes, M = 2.68, SD = 1.13,
F(1, 36) = 35.46, p , .001, h2 = .50.More relevant to testing the assumptions, the two-way interaction between out-
come valence and social distance was significant, F(1, 76) = 5.97, p , .05, h2 = .04;see Figure 3. Planned contrasts (with order controlled) were conducted to explore
the nature of the two-way interaction between social distance and outcome valence.The results indicated that participants generated more positive outcomes of refrain-
ing from smoking when asked to do so from the perspective of an average under-graduate student (M = 3.49, SE = .24) than when instructed to do so from their own
perspective, M = 2.82, SE = .23, F(1, 76) = 4.09, p , .05, h2 = .05. This finding isconsistent with the assumption that mental salience of positive outcomes of anaction increases as social distance increases. However, the number of negative out-
comes of smoking generated was not affected by social distance (distant: M = 4.03,SE = .31; proximal: M = 3.99, SE = .28, F , 1). The interaction between social
distance and outcome valence was not dependent upon order as the three-wayinteraction was not significant (p . .10). Overall, the findings appear to mirror
the results of Experiment 1, where it was found that the persuasiveness of a gainframe increased as social distance increased, whereas the persuasiveness of a loss
frame was not affected by social distance.Two key assumptions underlying H2 are that mental salience of societal out-
comes of an action increases as social distance increases; whereas mental salience of
individual outcomes of an action decreases as social distance increases. To test theseassumptions, a Social Distance (proximal or distant) 3 Order (positive first or
negative first) 3 Outcome Orientation (individual or societal) mixed ANOVAwas performed on the number of listed outcomes, with social distance and order
as between-subjects factors, and outcome orientation as a within-subjects factor.The results revealed only a significant main effect of outcome orientation,
F(1, 76) = 201.98, p , .001, h2 = .72. The number of individual outcomes (M = 6.40,
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 505
SD = 3.03) was considerably greater than that of societal outcomes (M = .80, SD = 1.19).
The expected interaction between social distance and outcome orientation was notsignificant. In the case of testing a priori assumptions, the omnibus interaction test
may be skipped and planned contrasts, which compare particular cells within a designand tend to have a greater statistical power, may be used to test the hypotheses(Keppel, 1991). Thus, even though the omnibus interaction was not significant,
planned contrasts were conducted to see if the empirical data were consistent withthe a priori assumptions. The analyses revealed that participants generated more
societal outcomes when asked to do so from the perspective of an average under-graduate student (M = 1.14, SD = 1.44) than when instructed to do so from their
own perspective, M = .52, SD = .87, F(1, 76) = 6.09, p , .05, h2 = .05; see Figure 4.This finding is consistent with the assumption that mental salience of societal out-
comes of an action increases as social distance increases. Apparently, the number ofindividual outcomes generated was apparently not affected by social distance (dis-
tant: M = 6.33, SD = 3.52; proximal: M = 6.45, SD = 2.60, F , 1). Overall, thefindings appear to mirror the results of Experiment 2, where it was found that thepersuasiveness of a societal frame increased as social distance increased, whereas
the persuasiveness of an individual frame was not affected by social distance.
Discussion
The key assumptions underlying the hypotheses were tested in Experiment 3. Two
assumptions that involve high-level construals were supported. Specifically, the resultsdemonstrated that mental salience of high-level construals, such as positive outcomes
and societal outcomes of an action, increases as social distance increases. These
2.8
3
3.2
3.4
3.6
3.8
4
4.2
Proximal DistantSocial Distance
Positive OutcomesNegative Outcomes
Figure 3 The number of reported outcomes as a function of social distance and outcome
valence (positive outcomes of refraining from smoking vs. negative outcomes of smoking).
Social Distance, Framing, and Judgment X. Nan
506 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
findings support the construal level explanation for findings obtained in Experiment 1,where it was shown that the persuasiveness of a gain frame increased as social distance
increased, and for results observed in Experiment 2, where it was found that thepersuasiveness of a societal frame increased as social distance increased.
The two assumptions involving low-level construals were not supported. Inaddition, Experiments 1 and 2 did not find any differential persuasive effect of a lossframe and an individual frame at varying levels of social distance. Although low
statistical power was a potential factor contributing to the null effects, the consis-tency in the findings obtained in three experiments nonetheless pointed to the
possibility that the salience of low-level construals is not as affected by psychologicaldistance, compared to the salience of high-level construals. Indeed, Pennington and
Roese’s (2003) studies have demonstrated just this point.
General discussion
Overall findings
CLT has been central in the investigation of the effects of temporal distance onperceptions and judgments. However, it has been utilized rarely to gain a better
understanding of the perceptual and judgmental influences of social distance,another common form of psychological distance. This research relates the study of
social distance to an important area of research in communication, namely framingin persuasive messages.
Two types of framing were the focus of this research: gain–loss framing andsocietal–individual framing. Drawing upon previous work, this article proposed that
a gain frame and a societal frame both constitute high-level construals and,
0
1
2
3
4
5
6
Proximal DistantSocial Distance
Prop
ortio
n of
Rep
orte
d O
utco
mes
7
Societal OutcomesIndividual Outcomes
Figure 4 The number of reported outcomes as a function of social distance and outcome
orientation (societal outcomes vs. individual outcomes).
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 507
according to CLT, will be more influential in judgments made for socially distantentities than for socially proximal entities. Both hypotheses were supported by
empirical evidence. Experiment 1 found that a gain frame was more persuasive whenpeople make judgments for their best friend than for an average undergraduate
student. Experiment 2 showed that a societal frame was more persuasive whenpeople make judgments for an average undergraduate student than for themselves.Using a thought-listing technique, Experiment 3 demonstrated that mental salience
of positive and societal outcomes of an action increased as social distance increased,which were assumptions underlying H1a and H2a. Therefore, these findings support
the viability of a construal level explanation for the differential persuasive effects ofgain frames and societal frames across varying levels of social distance.
Drawing upon previous research, I also argued that loss frames and individualframes constitute low-level construals and, thus, should be more persuasive when
judgments are made for socially proximal entities than for socially distant entities.These predictions were not supported by empirical data. Neither Experiment 1 norExperiment 2 found any systematic difference in the persuasiveness of the frames
across the two social distance conditions. Experiment 3, which used a thought-listingtechnique to tap into the salience of negative and individual outcomes of an action,
similarly did not produce any difference in the mental salience of the outcomes atvarying levels of social distance.
Theoretical implications
Overall, findings from the three experiments are consistent with CLT. The findingsalso indicate the need for a more refined theory. CLT research predicts that the
salience of high-level construals will increase, whereas the salience of low-level con-struals will decrease as psychological distance increases. However, the current studiesas well as some previous studies (e.g., Pennington & Roese, 2003) found no signifi-
cant difference in the salience of low-level construals across different levels of psy-chological distance. One possible explanation is that low-level construals, being
concrete and contextualized, are less sensitive to changes in psychological distancecompared to high-level construals. Thus, whether or not psychological distance
exerts an influence on the salience of low-level construals may be dependent uponthe operationalizations of psychological distance and construal level. Differences in
mental salience may emerge when there is sufficient distance and when the construallevel is sufficiently low (i.e., when the construals are sufficiently concrete anddetailed). Although answers to this question lie beyond the scope of this paper, more
research is needed to explore the possibilities raised here.More important to communication research, this inquiry holds relevant impli-
cations for framing studies. Over the years, gain–loss framing has received consider-able research attention, particularly in the area of health communication. The focus
has been on the relative persuasiveness of a gain versus loss frame. It is important tonote that most previous research has centered on the impact of gain–loss framing on
personal judgments (i.e., judgments made for selves). This research introduces the
Social Distance, Framing, and Judgment X. Nan
508 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
notion of psychological distance to the study of gain–loss framing, suggesting that aspsychological distance increases (e.g., when judgments are made for others rather
than selves), the impact of a gain frame (a high-level construal) will increase, whereasthe effect of a loss frame (a low-level construal) will decrease. This implies that the
observed relative effect of a gain versus loss frame when only personal judgments areinvolved could be amplified, diminished, or reversed in judgments made for sociallydistant entities. Specifically, if a gain frame is found to be more persuasive than a loss
frame in personal judgments, the effect could be amplified in judgments made forothers. If a loss frame is found to be more persuasive in personal judgments, the
effect could be diminished or even reversed in judgments made for socially distantentities.
Research on societal–individual framing has not been as active. This researchhighlights the possibility that a persuasive message can be framed either at a societal
level, focusing on the implications of compliance/noncompliance for the society ingeneral, or at an individual level, emphasizing personal consequences. More import-antly, this research suggests that a societal frame differs from an individual frame in
terms of construal level. As demonstrated here, a societal frame is generally associ-ated with a high-level construal about the advocated issue and its effect on persua-
sion increases as psychological distance increases. An individual frame is associatedwith a low-level construal and its effect remains constant at different levels of psy-
chological distance.At a broader level, this research complements previous investigation of the
perceptual impact of social distance. The field of communication has had a longtradition of studying the effects of social distance on perceptions. Research on the
third-person effect (i.e., the general tendency for people to believe that media con-tent has a greater impact on others than on themselves) epitomizes this tradition(Davison, 1983; McLeod, Detenber, & Eveland, 2001; Perloff 1989; Price, Tewksbury, &
Huang, 1998). Over the years, researchers have explored different explanations forthe third-person effect. The proposed explanations generally fall into one of two cat-
egories. Some are motivation-based explanations, suggesting that the third-personeffect results from an individual’s motivation to enhance self-esteem (Gunther &
Thorson, 1992; Meirick, 2005; Perloff, 2002). Others are cognition-based explana-tions that emphasize the role of cognitions in the third-person effect (Eveland,
Nathanson, Detenber, & McLeod, 1999; Gunther 1991; McLeod, Eveland, & Nathanson,1997). Gunther imported the theory of fundamental attribution error from socialpsychology to explain the third-person effect. Gunther argued that when judging
message impact on others, observers will underestimate the effect of situationalfactors and perceive more opinion change of those others, but in judging themselves,
observers will estimate modest opinion change due to a greater awareness of situa-tional factors.
By approaching the perceptual and judgmental effects of social distance from theperspective of CLT, this research offers a unique angle to view the third-person effect.
From the perspective of CLT, the third-person effect may be partly driven by the
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 509
different construal level of people’s mental representations for their own versusothers’ behaviors. According to CLT, people tend to form concrete and contextual-
ized representations (i.e., low-level construals) for their own behaviors. However, forothers’ behaviors, the mental representations tend to be abstract and decontextual-
ized (i.e., high-level construals). As such, people may perceive more complexity intheir own behaviors versus others’ and believe that their own behaviors are deter-mined by more factors. This could result in a less perceived impact of media content
on selves than on others. The construal level account of the third-person effect is inessence consistent with the explanation based on the fundamental attribution error.
It is more aligned with cognition-based explanations of the third-person effect thanto motivation-based explanations.
Implications for health campaigns
Although most health campaigns are targeted directly at the group of people whosehealth beliefs, attitudes, or behaviors are to be changed, there has been a growingtrend for health messages to indirectly target influence groups (e.g., friends, parents).
The purpose of such campaigns is to urge these social groups to exert a positiveinfluence on others, who are the ultimate target audience of the messages. For
example, the recent ‘‘Friends Watch Out for Friends’’ campaign launched on thecampus of Michigan State University aimed at persuading students to prevent their
friends from engaging in high-risk drinking. Despite the amount of research onhealth communication, there has been scant research on this campaign approach.
To what extent are these campaigns effective in motivating people to help others?What message strategies should we use to make a persuasive appeal?
This research indicates that the same message strategy may be differentially effec-tive depending on whether the message recipients act as the ultimate target audience orinfluence groups. When asked to make judgments for others (i.e., acting as the influ-
ence groups), people are more persuaded by high-level construals of the advocatedissue such as a gain frame and a societal frame, compared to when asked to make
judgments for themselves (i.e., acting as the ultimate target audience). Low-level con-struals such as a loss frame and an individual frame are less susceptible to role changes.
Limitations and future research
Certain limitations of this research are to be acknowledged. The first limitationpertains to the within-subjects manipulation of social distance in the first two experi-ments. By asking participants similar questions twice, albeit with different reference
groups (i.e., best friend vs. average student, or self vs. average student), the proceduremay have artificially increased the response differences. Associated with the within-
subjects design is the issue of possible question order effect, which was not explicitlyaddressed in the first two experiments. Future studies should replicate the studies
using a between-subjects design.A second limitation of this research is the limited number of dependent variables
relating to message effects. Although issue judgment as operationalized in this
Social Distance, Framing, and Judgment X. Nan
510 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
research is an important indicator of persuasion, there are a number of other depen-dent variables that are potentially interesting to both academics and practitioners.
These include message comprehension, knowledge gain and recall, and affectiveand cognitive responses. Of course, to systematically investigate the joint effects of
framing and social distance on these dependent variables, additional theory-basedhypotheses need to be proposed.
Another limitation associated with the first two experiments is the lack of manipu-
lation checks. One could check the success of the framing manipulations by askingparticipants directly about their perceptions about the message features. For
instance, participants in the gain frame condition would be asked to indicate theextent to which they believe the persuasive message focuses on the positive outcomes
of performing the advocated behavior. Results of such manipulation checks areusually very predictable (e.g., Block & Keller, 1995; Maheswaran & Meyers-Levy,
1990). Although desirable, it may be less important to check the manipulation ofsocial distance. Conforming to the instruction of focusing on oneself or anotherperson when making judgments should not be difficult for participants.
In addition to those highlighted above, this research opens up new avenues forfuture research. First, more systematic research on how social distance influences the
construal level of mental representations is needed. The key research question is howpeople mentally represent information pertaining to selves and others and whether
the two types of representations differ in terms of construal level (e.g., concretenessor abstractness). Second, and more relevant to strategic communication research,
future research should continue to test the idea that high-level construals are moresalient and, thus, important in making judgments for socially distant versus proximal
entities, whereas the reverse is true for low-level construals. In doing so, futurestudies should either replicate the current studies or utilize different operationaliza-tions of construal level (e.g., another message feature).
Conclusion
In addition to those highlighted above, this research opens up a few avenues forfuture research. First, more systematic research on how social distance influences the
construal level of mental representations is needed. The key research question is howpeople mentally represent information pertaining to selves and others and whether
the two types of representations differ in terms of construal level (e.g., concretenessor abstractness). Second, more relevant to strategic communication research, futureresearch can continue to test the idea that high-level construals are more salient and
thus important in making judgments for socially distant versus proximal entities,whereas the reverse is true for low-level construals. In doing so, future studies can
either replicate the current studies or utilize different operationalizations of con-strual level (e.g., another message feature). Understanding how social distance influ-
ences people’s perceptions and the way they make judgments is an importantscientific inquiry. By approaching this issue uniquely from a construal level perspec-
tive and relating it to the study of framing effects, this research hopefully serves as
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 511
a springboard for future inquiries in the realm of health communication or strategiccommunication in general.
References
Block, L. G., & Keller, P. A. (1995). When to accentuate the negative: The effects of perceived
efficacy and message framing on intentions to perform a health-related behavior. Journal
of Marketing Research, 32, 192–203.
Davis, J. J. (1995). The effects of message framing on response to environmental commu-
nications. Journalism & Mass Communication Quarterly, 72, 285–299.
Davison, W. P. (1983). The third-person effect in communication. Public Opinion Quarterly,
47, 1–15.
Ebert, J. E. J. (2005). Self-sympathy in the short-term: Self-other differences in long-term
benefits and short-term costs. Advances in Consumer Research, 32, 182–183.
Eveland, W. P., Jr., Nathanson, A. I., Detenber, B. H., & McLeod, D. M. (1999). Rethinking
the social distance corollary: Perceived likelihood of exposure and the third-person
perception. Communication Research, 26, 275–302.
Eyal, T., Liberman, N., & Trope, Y. (2004). The pros and cons of temporally near and distant
action. Journal of Personality and Social Psychology, 86, 781–795.
Fiedler, K., Semin, G. R., Finkenauer, C., & Berkel, I. (1995). Actor-observer bias in close
relationships: The role of self-knowledge and self-related language. Personality and
Social Psychology Bulletin, 21, 525–538.
Forster, J., Friedman, R. S., & Liberman, N. (2004). Temporal construal effects on abstract and
concrete thinking: Consequences for insight and creative cognition. Journal of Personality
and Social Psychology, 87, 177–189.
Gunther, A. C. (1991). What we think others think: Cause and consequence in the
third-person effect. Communication Research, 18, 355–372.
Gunther A. C., & Thorson, E. (1992). Perceived persuasive effects of product commercials and
public service announcements: Third-person effects in new domains. Communication
Research, 19, 574–596.
Jones, E. E., Wood, G. C., & Quattrone, G. A. (1981). Perceived variability of personal
characteristics in ingroups and outgroups: The role of knowledge and evaluation.
Personality and Social Psychology Bulletin, 7, 523–528.
Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk.
Econometrica, 47, 263–291.
Keppel, G. (1991). Design and analysis: A researcher’s handbook (3rd ed.). Englewood Cliffs,
NJ: Prentice Hall.
Levin, I. P., Schneider, S. L., & Gaeth, G. J. (1998). All frames are not created equal: A typology
and critical analysis of framing effects. Organizational Behavior and Human Decision
Processes, 76, 149–188.
Lewin, K. (1951). Field theory in social science. New York: Harper.
Liberman, N., & Trope, Y. (1998). The role of feasibility and desirability considerations in
near and distant future decisions: A test of temporal construal theory. Journal of
Personality and Social Psychology, 75, 5–18.
Social Distance, Framing, and Judgment X. Nan
512 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
Linville, P. W., Fischer, G. W., & Yoon, C. (1996). Perceived covariation among the features of
ingroup and outgroup members: The outgroup covariation effect. Journal of Personality
and Social Psychology, 70, 421–436.
Maheswaran, D., & Meyers-Levy, J. (1990). The Influence of message framing and issue
involvement. Journal of Marketing Research, 27, 361–367.
McLeod, D. M., Detenber, B. H., & Eveland, W. P., Jr (2001). Behind the third-person effect:
Differentiating perceptual processes for self and other. Journal of Communication, 51, 678–685.
McLeod, D. M., Eveland, W. P., Jr., & Nathanson, A. I. (1997). Support for censorship of
violent and misogynic rap lyrics: An analysis of the third-person effect. Communication
Research, 24, 153–174.
Meirick, P. C. (2005). Self-enhancement motivation as a third variable in the relationship
between first- and third-person effects. International Journal of Public Opinion Research,
17, 473–483.
Meyerowitz, B. E., & Chaiken, S. (1987). The effect of message framing on breast
self-examination attitudes, intentions, and behavior. Journal of Personality and
Social Psychology, 52, 500–510.
Mitchell, M. M. (2001). Risk, Threat, and information seeking about genital herpes: The
effects of mood and message framing. Communication Studies, 52, 141–152.
O’Keefe, D. J., & Jensen, J. D. (2006). The advantages of compliance or the disadvantages of
noncompliance? A meta-analytic review of the relative persuasive effectiveness of
gain-framed and loss-framed messages. Communication Yearbook, 30, 1–43.
O’Keefe, D. J., & Jensen, J. D. (in press-a). Do loss-framed persuasive messages engender
greater message processing than do gain-framed messages? A meta-analytic review.
Communication Studies.
O’Keefe, D. J., & Jensen, J. D. (in press-b). The relative persuasiveness of gain-framed and
loss-framed messages for encouraging disease prevention behaviors: A meta-analytic
review. Journal of Health Communication.
Pennington, G. I., & Roese, N. J. (2003). Regulatory focus and temporal distance. Journal
of Experimental Social Psychology, 39, 563–576.
Perloff, R. M. (1989). Ego-involvement and the third person effect of televised news coverage.
Communication Research, 16, 236–262.
Perloff, R. M. (2002). The third-person effect. In B. Jennings & D. Zillman (Eds.), Media
effects: Advances in theory and research (2nd ed., pp. 489–506). Mahwah, NJ: Erlbaum.
Price, V., Tewksbury, D., & Huang, L. (1998). Third person effects on publication of
a Holocaust-denial advertisement. Journal of Communication, 48(2), 3–26.
Rothman, A. J., Bartels, R. D., Wlaschin, J., & Salovey, P. (2006). The strategic use of gain- and
loss-framed messages to promote healthy behavior: How theory can inform practice.
Journal of Communication, 56, S202–S220.
Rothman, A. J., Salovey, P., Antone, C., Keough, K., & Martin, C. (1993). The influence of
message framing on health behavior. Journal of Experimental Social Psychology, 29,
408–433.
Shah, D. V., Faber, R. J., & Youn, S. (1999). Susceptibility and severity: Perceptual dimensions
underlying the third-person effect. Communication Research, 26, 240–267.
Shah, D. V., Kwak, N., Schmierbach, M., & Zubric, J. (2004). The interplay of news frames
on cognitive complexity. Human Communication Research, 30, 102–120.
Trope, Y., & Liberman, N. (2003). Temporal construal. Psychological Review, 110, 403–421.
X. Nan Social Distance, Framing, and Judgment
Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association 513
Umphrey, L. R. (2003). The effects of message framing and message processing on testicular
self-examination attitudes and perceived susceptibility. Communication Research Reports,
20(2), 97–105.
Weinstein, N. D. (1987). Unrealistic optimism about susceptibility to health problems:
Conclusions from a community-wide sample. Journal of Behavioral Medicine, 10,
481–498.
Ziamou, P., & Veryzer, R. W. (2005). The influence of temporal distance on consumer
preferences for technology-based innovations. Journal of Product Innovation Management,
22, 336–346.
Social Distance, Framing, and Judgment X. Nan
514 Human Communication Research 33 (2007) 489–514 ª 2007 International Communication Association
Distance sociale, cadrage et jugement: une perspective de niveaux de
représentation
Xiaoli Nan
University of Wisconsin-Madison
Résumé
Mobilisant la théorie des niveaux de représentation, cette recherche étudie l’influence de la
distance sociale sur les réponses de personnes à des messages persuasifs. L’expérience 1 (N =
133) démontre que l’impact persuasif d’un cadre de gain est plus fort lorsque les gens exercent
leur jugement pour des entités distantes (p. ex. d’autres personnes) par opposition à des entités
proximales (p. ex. soi-même). En revanche, l’impact persuasif d’un cadre de perte demeure le
même à travers différents niveaux de distance sociale. L’expérience 2 (N = 135) démontre que la
force de persuasion d’un cadre sociétal est plus forte lorsque les gens exercent leur jugement
pour des entités socialement distantes plutôt que pour des entités proximales, tandis que la force
de persuasion d’un cadre individuel n’est pas affectée par la distance sociale. L’expérience 3 (N
= 80) offre une preuve de ce que la prégnance mentale des conséquences positives et sociétales
d’une action augmente comme la distance sociale augmente, alors que la prégnance mentale des
conséquences négatives et individuelles demeure la même à travers différents niveaux de
distance sociale.
Soziale Distanz, Framing und Urteilen: Eine Bedeutungsebenenperspektive
Vorliegender Beitrag untersucht den Einfluss von sozialer Distanz auf Reaktionen zu
persuasiven Botschaften unter Rückgriff auf eine Bedeutungsebenentheorie. Experiment 1
(N=133) zeigte, dass der persuasive Einfluss eines Zugewinn-Frames stärker wird, wenn
Personen Urteile über sozial entfernte (z.B. andere) vs. sozial nahe Instanzen (z.B. selbst) fällen.
Auf der anderen Seite zeigte sich, dass der persuasive Einfluss eines Verlust-Frames keine
unterschiedlichen Ergebnisse über verschiedene Ebenen sozialer Distanz hervorbringt. In
Experiment 2 (N=135) konnte gezeigt werden, dass die Überzeugungskraft eines sozialen
Frames stärker wird, wenn Personen Urteile für sozial entfernte vs. nahe Instanzen machen,
während die Überzeugungskraft eines individuellen Frames von der sozialen Distanz
unbeeinflusst bleibt. Experiment 3 (N=80) liefert Belege dafür, dass sich die mentale Salienz von
positiven und gesellschaftlichen Folgen einer Handlung vergrößert, wenn die soziale Distanz
wächst, während die mentale Salienz von negativen und individuellen Folgen über verschiedene
Ebenen der sozialen Distanz unverändert bleibt.
La Distancia Social, el Framing, y el Juicio: Una Perspectiva a Nivel Interpretativo
Xiaoli Nan
University of Wisconsin-Madison
Resumen
Basada en la teoría de Nivel Interpretativo, esta investigación explora la influencia de la
distancia social sobre las respuestas individuales a los mensajes persuasivos. El
experimento 1 (N = 133) demuestra que el impacto persuasivo de un encuadre de
ganancia se vuelve más poderoso cuando la gente emite juicios sobre entidades
socialmente distantes (a saber, otros) versus entidades próximas (a saber, si mismos). Por
otro lado, el impacto persuasivo de un encuadre de pérdida se mantiene igual a través de
los distintos niveles de distancia social. El experimento 2 (N = 135) muestra que el nivel
de persuasión de un encuadre social se vuelve más poderoso cuando la gente emite
juicios sobre entidades socialmente distantes versus entidades próximas, mientras que el
nivel de persuasión de un encuadre individual no es afectado por la distancia social. El
experimento 3 (N = 80) provee evidencia que la notabilidad mental de resultados
positivos y sociales de una acción se incrementan a medida que la distancia social
aumenta, mientras que la notabilidad mental de los resultados negativos e individuales se
mantienen a través de los distintos niveles de distancia social.
社会距离,框架和判断:解释层次观点
Xiaoli Nan
威斯康辛大学
[摘要]
本文以解释层次理论为基础,研究社会距离给个人对说服性信息
的影响。实验一(N=133)表明相对于邻近的个体(如自我),人们对有
社会距离的个体(如别人)作判断时得益框架的说服力更强。然而,
丧失框架在所有社会距离的作用则不变;实验二(N=135)表明,相
对于邻近的个体,人们对有社会距离的个体判断时社会框架的说服力
更强,而个体框架则不受社会距离影响;实验三(N=80)的证据显
示:随着社会距离的增加行为的正面和社会结果的思想显著度也增
加,然而,负面的和个人的结果其思想显著度在不同社会距离下则无
区别。
사회적 거리감, 프레이밍, 그리고 판단: 해석 수준 전망에 관한 연구
Xiaoli Nan
University of Wisconsin-Madison
요약
해석수준이론에 근거하여, 본 연구는 설득적 메시지들에 대한 개인들의 반응에 있어
사회적 거리감의 영향을 연구한 것이다. 실험 1 (N = 133)은 이익 프레임의 설득적
영향은 사람들이 사회적 거리감 (예컨대 타자들)과 가까운 실제들 (예를들어 자기들)에
대한 판단을 내릴때 더욱 강해지는 것으로 나타났다. 한편, 손실프레임의 설득적 영향은
사회적 거리감의 여러 다른 차원들에 걸쳐 같은 정도를 보이는 것으로 나타났다. 실험 2
(N = 135)에서는 사회적 프레임의 설득력은 사람들이 사회적인 거리감대 인접한
대상들에 대한 판단을 내릴때 더욱 강하게 나타나는 반면, 개인적 프레임의 설득력은
사회적 거리감에 의해 영향을 받지 않는다는 것을 보여주고 있다. 실험 3 (N = 80)은
어떤 행위의 긍적적이고 사회적인 결과들의 정신적 현명함은 사회적 거리감이
증가함에 따라 커지는 반면, 부정적이고 개인적인 결과들의 정신적 정신적 현명함은
사회적 거리감의 전반적인 수준에 있어 같은 상태로 남아있음을 보여주고 있다.